Ivan Maynart Tavares

Bayesian Machine Learning Classifiers for Combining Structural and Functional Measurements to Classify Healthy and Glaucomatous Eyes

PURPOSE To determine whether combining structural (optical coherence tomography, OCT) and functional (standard automated perimetry, SAP) measurements as input for machine learning classifiers (MLCs; relevance vector machine, RVM; and subspace mixture of Gaussians, SSMoG) improves diagnostic accuracy for detecting glaucomatous eyes compared with using each measurement method alone. METHODS Sixty-nine eyes of 69 healthy control subjects (average age, 62.0, SD 9.7 years; visual field mean deviation [MD], -0.70, SD 1.41 dB) and 156 eyes of 156 patients with glaucoma (average age, 66.4, SD 10.2 years; visual field MD, -3.12, SD 3.43 dB) were imaged with OCT (Stratus OCT, Carl Zeiss Meditec, Inc., Dublin, CA) and tested with SAP (Humphrey Field Analyzer II with Swedish Interactive Thresholding Algorithm, SITA; Carl Zeiss Meditec, Inc.) within 3 months of each other. RVM and SSMoG MLCs were trained and tested on OCT-determined RNFL thickness measurements from 32 sectors ( approximately 11.25 degrees each) obtained in the circumpapillary area under the instrument-defined measurement ellipse and SAP pattern deviation values from 52 points from the 24-2 grid, independently and in combination. Tenfold cross-validation was used to train and test classifiers on unique subsets of the full 225-eye data set, and areas under the receiver operating characteristic curve (AUROC) for the classification of eyes in the test set were generated. AUROC results from classifiers trained on OCT and SAP alone and those trained on OCT and SAP in combination were compared. In addition, these results were compared to currently available OCT measurements (mean retinal nerve fiber layer [RNFL] thickness, inferior RNFL thickness, and superior RNFL thickness) and SAP indices (MD and pattern standard deviation [PSD]). RESULTS The AUROCs for RVM trained on OCT parameters alone, SAP parameters alone and OCT and SAP parameters combined were 0.809, 0.815, and 0.845, respectively. The AUROCs for SSMoG trained on OCT parameters alone, SAP parameters alone, and OCT and SAP parameters combined were 0.817, 0.841, and 0.869, respectively. Combining techniques using both RVM and SSMoG significantly improved on MLC analysis of OCT, but not SAP, measurements alone. Classification performance using RVM and SSMoG was statistically similar. CONCLUSIONS RVM and SSMoG Bayesian MLCs trained on OCT and SAP data can successfully discriminate between healthy and early glaucomatous eyes. Combining OCT and SAP measurements using RVM and SSMoG increased diagnostic performance marginally compared with MLC analysis of data obtained using each technology alone.

The retinal nerve fiber layer of patients with neuromyelitis optica and chronic relapsing optic neuritis is more severely damaged than patients with multiple sclerosis.

Backgroud: To compare the retinal nerve fiber layer (RNFL) in eyes of patients with relapsing remitting multiple sclerosis (RRMS), neuromyelitis optica (NMO) and chronic relapsing inflammatory optic neuritis (CRION). Methods: Evaluation of 62 patients with RRMS, NMO, and CRION in a cross-sectional study with spectral domain optical coherence tomography. Results: A total of 124 eyes were evaluated (96 RRMS, 18 NMO, and 10 CRION). Frequency of optic neuritis for each disease was: 34% for RRMS, 84% for NMO, and 100% for CRION. Visual acuity and RNFL thickness were significantly worse in NMO and CRION eyes than in RRMS, but there were no differences between NMO and CRION eyes. A RNFL of 41 &mgr;m was 100% specific for optic neuritis associated with NMO and CRION when compared to RRMS. Conclusion: This study established RNFL values to differentiate optic neuritis of RRMS from NMO and CRION. Although similarities observed between NMO and CRION eyes might suggest that they are within the same disease spectrum, it is still recommended that these 2 conditions be differentiated on clinical grounds. Optical coherence tomography serves as an additional diagnostic tool and can be used to monitor disease progression.

Relationship between retinal nerve fiber layer and visual field function in human immunodeficiency virus-infected patients without retinitis.

Purpose: To evaluate the spatial association between visual field (VF) sensitivity loss and retinal nerve fiber layer (RNFL) thinning in patients infected by the human immunodeficiency virus. Methods: Fifty-one eyes of 51 human immunodeficiency virus–infected patients and 22 eyes of 22 control subjects were enrolled. Patients were evaluated using the Fast RNFL scan strategy on Stratus OCT and the 24-2 full-threshold program on the Humphrey Matrix frequency doubling technology (FDT) perimeter. Associations between RNFL thickness and VF sensitivity were evaluated globally, in 12 clock-hour optical coherence tomography sectors and in 21 VF zones; linear and quadratic regression models were used in the statistical analysis. Results: The linear and quadratic regression associations between the FDT Matrix pattern standard deviation and the average RNFL thickness in human immunodeficiency virus–infected patients were r2 = 0.185 and r2 = 0.218 (P < 0.05), respectively. The correlation between the FDT Matrix mean deviation and the average RNFL thickness was not significant (P > 0.05). Stronger associations were found when regional RNFL thinning was compared with locally measured FDT Matrix pattern deviation, especially between nasal RNFL measurements and temporal VF zones, and between superior RNFL measurements and inferior VF zones. Conclusion: Retinal nerve fiber layer thinning was related to VF sensitivity loss in human immunodeficiency virus–infected patients and regional associations between optical coherence tomography and FDT Matrix sectors were stronger than the associations between global measurements.

Inconsistency of the published definition of ocular hypertension.

PurposeTo examine the variability and to evaluate the influence of the Ocular Hypertension Treatment Study (OHTS) on the criteria for published definition of ocular hypertension (OH). DesignSystematic literature review. MethodsAll articles about OH published between January 1995 and July 2005 from the 3 clinical journals of ophthalmology were identified by a literature search. They were reviewed to determine the criteria used to define OH, including intraocular pressure and central corneal thickness, visual field and optic disc assessment. ResultsThere were 133 studies published on OH during the period under study. As a goal of the current study was to evaluate the influence of the publication of the OHTS results on the definition of OH used in the literature, all reports from the OHTS and its ancillary studies (n=11) were excluded from the main analysis. After the exclusion of the OHTS reports, a total of 122 studies were used in the analysis. The intraocular pressure cut-off level used for OH diagnosis ranged from 18 to 26 mm Hg, with a mode of 22 mm Hg (49.2%). Only 13.1% of the articles reported central corneal thickness, and these were distributed evenly during the 10-year period under study. Visual field assessment was reported by most of the studies (110; 90.2%), but only 45% of them defined the criteria to classify a field as normal or glaucomatous. Further, only 37 of the 57 articles that reported the method of optic disc assessment described the criteria used to classify optic discs. ConclusionsThe criteria used to define OH during the past 10 years have been highly variable. This wide variation suggests the important need for a standardized definition of OH.

No changes in anatomical and functional glaucoma evaluation after trabeculectomy

PurposeTo assess the influence of glaucoma filtration surgery on anatomical and functional tests for glaucoma evaluation.MethodsTwenty-five eyes (25 patients) with primary open-angle glaucoma were evaluated, prospectively. Data were collected on vision acuity, intraocular pressure, standard automated perimetry, frequency doubling technology perimetry, scanning laser polarimetry (GDx) and confocal scanning laser ophthalmoscopy (HRT II) before and 3–6 months after surgery.ResultsMean (±SD) pre- and postoperative visual acuities (logMAR) were 0.28 (±0.18) and 0.30 (±0.17), respectively (P=0.346). In a mean time of 4.5 (±1.1) months after surgery, the mean preoperative intraocular pressure of 20.7 (±5.4) mmHg decreased to 11.04 (±2.52) mmHg (P<0.001). The results of the standard automated perimetry, frequency doubling technology perimetry, scanning laser polarimetry and confocal scanning laser ophthalmoscopy diagnostic methods revealed no significant difference (P>0.162) between pre and postoperative values and no significant correlation (P>0.296) between intraocular pressure reduction and value changes.ConclusionNo significant change on any test variable was detected after glaucoma filtration surgery. Trabeculectomy does not appear to influence standard automated perimetry, frequency doubling technology perimetry, scanning laser polarimetry and confocal scanning laser ophthalmoscopy (HRT II) results after a 4.5-month period of surgery in early to moderate glaucoma.

Hypotensive effect of juxtascleral administration of anecortave acetate in different types of glaucoma.

PurposeTo evaluate the intraocular pressure (IOP)-lowering effect of anecortave acetate, delivered by anterior juxtascleral depot injection, in eyes with various forms of glaucoma. MethodsA prospective, interventional case series was carried out. Twenty-eight uncontrolled glaucoma patients received a single injection of anecortave acetate (24 to 30 mg) in 1 selected eye under topical anesthesia. Postinjection assessments were scheduled at week 1 and months 1, 2, and 3. ResultsMean±SD age of patients was 58.2±18.6 years. Twelve patients had open-angle glaucoma and 16 had angle-closure glaucoma. Uveitic/steroid-induced glaucoma was the most frequent diagnosis (11 patients, 39.2%). Mean baseline IOP was 30.7±9.3 mm Hg. Mean IOP at week 1 and months 1, 2, and 3 were 21.3±6.1, 19.8±6.3, 20.9±7.3, and 21.7±6.8 mm Hg, respectively. Significant mean IOP reductions were observed at week 1 and months 1, 2, and 3 (29.3%, 33.8%, 30.1%, and 27.2%, respectively; P<0.001). ConclusionA single administration of anecortave acetate by anterior juxtascleral depot injection seems to result in a significant IOP reduction (compared with baseline) for at least 3 months in eyes with different types of glaucoma. No injection-related or drug-related serious adverse events were observed. Additional studies are required to better determine the efficacy, safety, and mechanism of action of this alternative treatment.

Normal-tension glaucoma: an update

Normal-tension glaucoma (NTG) is a progressive optic neuropathy with intraocular pressure (IOP) within the statistically normal range (≤21 mmHg). The prevalence of NTG varies widely among different population studies, being the most prevalent open-angle glaucoma subtype in some reports. The etiology of NTG possibly is multifactorial and still not well defined. Alternative treatments have been proposed based on pathogenesis details. However, in clinical practice, adequate reduction of IOP remains the keystone of managing patients with NTG. We review the pathogenesis of NTG and the available therapies for this optic neuropathy.

Genotype-Phenotype Correlations in CYP1B1-Associated Primary Congenital Glaucoma Patients Representing Two Large Cohorts from India and Brazil.

Background Primary congenital glaucoma (PCG), occurs due to the developmental defects in the trabecular meshwork and anterior chamber angle in children. PCG exhibits genetic heterogeneity and the CYP1B1 gene has been widely implicated worldwide. Despite the diverse mutation spectra, the clinical implications of these mutations are yet unclear. The present study attempted to delineate the clinical profile of PCG in the background of CYP1B1 mutations from a large cohort of 901 subjects from India (n=601) and Brazil (n=300). Methods Genotype-phenotype correlations was undertaken on clinically well characterized PCG cases from India (n=301) and Brazil (n=150) to assess the contributions of CYP1B1 mutation on a set of demographic and clinical parameters. The demographic (gender, and history of consanguinity) and quantitative clinical (presenting intraocular pressure [IOP] and corneal diameter [CD]) parameters were considered as binary and continuous variables, respectively, for PCG patients in the background of the overall mutation spectra and also with respect to the prevalent mutations in India (R368H) and Brazil (4340delG). All these variables were fitted in a multivariate logistic regression model using the Akaike Information Criterion (AIC) to estimate the adjusted odds ratio (OR) using the R software (version 2.14.1). Results The overall mutation spectrum were similar across the Indian and Brazilian PCG cases, despite significantly higher number of homozygous mutations in the former (p=0.024) and compound heterozygous mutations in the later (p=0.012). A wide allelic heterogeneity was observed and only 6 mutations were infrequently shared between these two populations. The adjusted ORs for the binary (demographic) and continuous (clinical) variables did not indicate any susceptibility to the observed mutations (p>0.05). Conclusions The present study demonstrated a lack of genotype-phenotype correlation of the demographic and clinical traits to CYP1B1 mutations in PCG at presentation. However, the susceptibility of these mutations to the long-term progression of these traits are yet to be deciphered.

Glaucoma de pressão normal

Normal tension glaucoma is an optic neuropathy in which the retinal nerve fiber layer thickness is reduced, the optic nerve is pathologically excavated, and the visual field is disturbed, although intraocular pressure is classified as normal. Patients with normal tension glaucoma are a heterogeneous group in which many systemic pathologic conditions may be present, and presenting ocular clinical signs that are difficult to be identified in the early stages of the disease. Treatment is more challenging than in hypertensive glaucomas. The authors sought to review the pathophysiology and differential diagnosis of this condition assessing vascular, rheumatic, neurological and genetic aspects that should be studied, as well as its treatment. The sources of references are PubMed (MEDLINE), LILACS and Cochrane Library (CENTRAL) databases.

The disc damage likelihood scale: Diagnostic accuracy and correlations with cup-to-disc ratio, structural tests and standard automated perimetry

Our objective was to compare the diagnostic accuracies of and to determine the correlations between the disc damage likelihood scale (DDLS) and anatomical and functional tests used for glaucoma detection. A total of 54 healthy subjects (54 eyes) and 47 primary open-angle glaucoma patients (47 eyes) were included in this cross-sectional observational study. DDLS scores and cup-to-disc (C/D) ratios were evaluated. Subjects underwent standard automated perimetry (SAP), optic disc and retinal nerve fiber layer (RNFL) imaging with time and spectral-domain optical coherence tomography (TD and SD-OCT), Heidelberg Retina Tomograph (HRT II), and scanning laser polarimetry (GDx-VCC). Areas under the receiver operating characteristic curves (AROCs) for DDLS and diagnostic tests parameters were calculated. DDLS correlations (Spearmans rank) among these parameters were analyzed. Fifty-four eyes were healthy and 47 had glaucoma, including 16 preperimetric glaucoma. DDLS, vertical and horizontal C/D ratios had the largest AROCs (0.92, 0.94 and 0.91, respectively). DDLS diagnostic accuracy was better than the accuracies of HRT II parameters, TD and SD-OCT RNFL thicknesses, and SAP mean deviation (MD) index. There were no significant differences between the accuracies of the DDLS and the C/D ratios, TD-OCT vertical (0.89) and horizontal (0.86) C/D ratios, TD-OCT C/D area ratio (0.89), and GDx-VCC NFI (0.81). DDLS showed significant strong correlations with vertical (r = 0.79) and horizontal (0.74) C/D ratios, and with the parameters vertical C/D ratio and C/D area ratio from HRT II (both 0.77) and TD-OCT (0.75 and 0.72, respectively). DDLS had significant moderate correlations with most of the other structural measurements and SAP MD. The optic disc clinical evaluation with DDLS system and C/D ratio demonstrated excellent accuracy in distinguishing glaucomatous from healthy eyes. DDLS had moderate to strong correlations with most structural and functional parameters. These findings stress the importance of optic disc clinical examination to detect glaucoma in a clinical scenario.